This study investigated whether the parameters of character-classification performance obtained under conditions of limited practice, which are consistent with the characterization of memory search as a high speed internal exhaustive serial comparison process having an average rate of between 25 and 30 symbols per second, were altered after prolonged practice. Analysis of the data from 7 Ss who each performed under 3 positive set size conditions on each of 30 sessions revealed that throughout the course of the experiment: (a) the relationship between set size and response latency remained linear, (b) equality of slopes obtained from positive and negative trials was maintained, and (c) the estimate of scanning rate was stable, and its absolute value was almost identical to previous values obtained with limited practice.
An index of intermodal attention switching, A, derived from RT procedures with multiple stimuli, a single response, and 2 levels of stimulus certainty, was used to compare the performance of 12 normal and 12 schizophrenic Ss. Schizophrenics had significantly more difficulty switching attention between modalities than normals. It cannot be determined on the basis of these data whether intermodal switching time is longer for schizophrenic Ss, or whether the signal to switch attention is not picked up as reliably, or acted upon as reliably, by the schizophrenic Ss. This method is compared with a method which assumed that S has certain stimulus expectancies, determined by the stimulus presented on the immediately preceding trial, and the former is found to be more valid, sensitive, and efficient.
Recognition tasks in which the items have been assigned to one and only to one response category throughout an experiment typically reveal that the functions relating reaction time and positive set size are negatively accelerated and have slopes which decrease significantly with practice. The present experiments were designed to determine whether this practice effect is specific to the individual items practiced or if it depends, at least partially, on relations built up among the positive set items. The results permit rejection of both alternatives. In the extreme case tested, it was found that, after prolonged practice with the same items composing nested positive sets, these items could be replaced with unpracticed ones without producing any significant change in the effect of set size, reaction times, or error levels. Absence of change in overall reaction times in combination with stability ofslope values leads to the conclusion that not only do those practice effects which are related to the set-size effect transfer exceedingly well to unpracticed items, but practice effects on all other stages involved in item recognition also show almost total transfer to unpracticed items. It is concluded that, under the conditions of these experiments, the gain (through practice) in efficiency in processing the stimulus is independent of the stimulus.
The experimental method allowed false-positive as weIl as false-negative errors to be measured during visual search. Data were collected from eight Ss on four target-set size conditions on each of 25 days in order to determine whether the finding that search rate is independent of target-set size after practice might be attributable to a systematic variation in total error rate with set size. Even after practice, these Ss continued to search significantly faster for a single target than for target sets which were composed of five or more characters. Further, the incidence of total errors was much lower than in those studies in which search rate had been found to be independent of set size (i.e., parallel processing), even though these latter studies considered only the occurrence of false-negative errors. Additional analysis strongly suggests that: (1) attainment of parallel processing and a highly aceurate performance are incompatible, and (2) verbal instructions alone are not sufficient to control the accuracy levels.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.